MISRA 10.4 Array Index False Positive?

[Amir]

Hello,

I'm wondering if the following is a false positive:

bool Interpolate_S16(int16_t *y_out, int16_t x_in, const uint8_t col_size, int16_t *table)
{
   int16_t y1, y2, x1, x2, dx, dy;
   int16_t *arr_y = &table[0];
   int16_t *arr_x = &table[col_size];
   uint16_t y1_index;
   bool ret_val = false;

   if (x_in < arr_x[0])
   {
      *y_out = arr_y[0];
   }
   else if (x_in >= arr_x[col_size - 1])
   ...
   }

Let's ignore the comma's for now. The following line is giving me MISRA 10.4 violation because of col_size. If I change col_size to an int8_t, the violation goes away:

else if (x_in >= arr_x[col_size - 1])

I don't see any requirement that the index of an array must be the same type according to MISRA. I believe this is a false positive. unsigned ints should be allowed as array indexes for signed ints, and do not appear to be a violation of 10.4.

Regards,
Amir

[Georgiy Komarov]

Hello, Amir,
I think this is not a false positive, because 1 in this example have signed integer type, when col_size is unsigned uint8t. I suggest to use 1u instead:

else if (x_in >= arr_x[col_size - 1u])